Vaginal douching is a commonly practiced behavior and previous work in Baltimore found that 70 percent of women in the Baltimore STD clinics report current douching. Douching is associated with a number of adverse outcomes, including bacterial vaginosis (BV), cervical infection, pelvic inflammatory disease (PID), and increased risk for ectopic pregnancy. BV, in particular, is associated with an increased risk of pre-term delivery, low birth weight and HIV. Most studies evaluating the effect of douching on risk for BV have been limited in their evaluation of feminine hygiene behavior and have been largely cross-sectional, limiting direct causal inference. Little data are available on how douching affects the vaginal flora on either a short or long-term basis or why women douche. No studies to date have evaluated vaginal flora after discontinuation of douching. We propose a cross-over intervention trial of vaginal douching to evaluate the change in vaginal flora after cessation of douching. We will recruit 150 women who currently douche from posted flyers, the Bayview Medical Center Gynecology clinic and Baltimore City Health Department STD clinics. We will measure the vaginal flora from self-collected samples (with Nugent's criteria) twice per week during a one month douching period. Subjects will then be asked to stop douching, and we will measure vaginal flora twice weekly for two months. Women who return to their normal douching patterns at the conclusion of the study will also be evaluated for a second cross-over exposure sample. Participants will maintain daily sexual and hygiene diaries throughout the study. Generalized estimating equations will be used to assess the following: 1) the effect of vaginal douching cessation on the risk of BV; 2) the risk for BV among women who abstain from douching but then revert to their previous douching habits. Common vaginal douching practices, motivations and techniques will also be evaluated. Data from this research will begin to describe the natural history of BV and aid in the design of behavioral interventions for vaginal douching. [unreadable] [unreadable]